Reproduce “Attention Maps Know When to Think: Training-Free Difficulty-Aware Chain-of-Thought for Reasoning Segmentation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Chanseul Cho, Jaesung Jun, Dongheon Lee, Youngmin Ro
DA-CoT uses visual attention entropy to choose whether reasoning segmentation uses chain-of-thought (CoT) or direct prediction. A threshold set on unlabeled examples guides the choice; invalid output structure triggers one switch between modes. On ReasonSeg test, it achieves 64.6 generalized intersection-over-union (gIoU), versus 60.7 for All-CoT, taking 4.7 versus 9.9 seconds. Concentrated attention can still focus on a wrong region.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.